R E J E C T E D H U M A N C O R N E A L G R A F T S

I. C L I N I C A L STUDY

C.C. KOK-VAN ALPHEN, H.J.M. V ( )LKER-DIEBEN,

J .J .L. VAN DER WANT & G. V R E N S E N

fLeiden/Amsterdam, The Netherlands)

Keywords: Keratoplasty, Corneal graft, Rejection, Khodadoust line

ABSTRACT

As introduction to a light- and electron-microscopial examination of rejected human corneal grafts a summary is given of the clinical signs of rejection. It is extremely im- portant to recognize these signs in time and to initiate the correct therapy. Most fail- ures in corneal transplantation are due to rejection. Rejection reactions are often not re- cognized early enough. The differential diagnosis is difficult because so many factors can trigger off a rejection. As the diagnosis is so often missed it is a good thing to consider the clinical picture of graft rejection once more. Maumenee: 'most ophthal- mologists do not recognize the signs of early graft failure and usually do not refer the patient back for therapy until considerable oedema has developed from endothelial destruction' (1962). In the cases with a good prognosis the rejection percentage is + 12%; in vascularized corneas and complicated corneal transplantations as high as 75% (Polack, 1977).

COURSE OF THE IMMUNOLOGICAL REACTION

L y m p h o c y t e s f rom the graft migrate like inquisi t ive scouts and come into

contac t wi th the recipient ' s lymphocy tes . The ' e n e m y ' is recognized by the

recipient and ' t roops ' are mobi l ized to at tack the graft. In the first mon ths

af ter the t ransplanta t ion the wound in Descemet ' s m e m b r a n e is no t comple-

te ly closed and the 'killer-cells ' can reach the endothe l ium of the graft

th rough this wound ( Inomata , 1970; Polack, 1973). This is the quickest

route and for this reason the first mon ths are mos t critical. Re jec t ion begins

most f requent ly at the edge of the graft (Fig. 1), and the nex t most f requent

form begins in the endothel ium. There is also a diffuse form with inf i l t ra t ion

of the s t roma and vascularisation; in this form the graft can become com-

ple te ly opaque in 3 - 4 days (Fig. 2). When the interval be tween the trans-

p lan ta t ion and the re jec t ion is longer the immune -compe ten t cells are main-

l y derived f rom the urea . In these cases the endo the l ium of the graft is at-

tacked f rom all sides. In the type of re ject ion which begins locally the Kho- dadoust line can be seen, a line in the endo the l ium formed by the meet ing

of the hosti le endothel ia l cells. This line is regarded as proof that a re ject ion

is taking place (Fig. 3). In animal exper iments , where a (local) re ject ion is

Documenta Ophthalmologica 50, 275-282 (1981). 00124486/81/0502-0275 $ 1.20. �9 Dr. W. Junk B.V. Publishers, The Hague. Printed in The Netherlands.

275

Fig. 1. Rejection begins at the edge.

Fig. 2. Diffuse, rapid rejection.

expressely induced, this line is usuaUy seen. In clinical practice this is not always the case because the rejection often occurs diffusely. The urea is

276

Fig. 3. Khodadoust line.

Fig. 4. Descemet precipitates.

involved in all forms of rejection. Descemet precipitates (Fig. 4) and a flair in the anterior chamber are seen. Occasionally the rejection begins in the

277

epithelium. The epithelium of the graft is replaced within a few days by that

of the recipient. It is difficult to see how a fatal interaction of cells can then

occur at a later date, nevertheless we know of cases where the rejection

clearly began in the epithelium. Every external noxa can give rise to a rejec-

t ion (Fig. 5). Traumata caused by the operation, such as anterior synechiae,

Operative trauma Recurrence of herpes Bacterial or fungal infections Other disease: heart infarct, operations, etc. Emotions

Fig. 5. Factors which can trigger off rejection.

a tight suture, irritation from a knot, loose sutures, can trigger off a rejec- tion at any time after the transplantation. A recurrence of herpes or another

infection can be associated with rejection. In these cases the corticosteroid

medication must on no account be stopped. The rejection can also be ac-

companied by glaucoma. Here again the corticosteroids must be continued,

but preparations with a less pressure-raising effect can be given, such as

prednisolon 2.5% or F.M.L. Other diseases, such as a heart infarct, operation, etc., can also cause a

rejection. The likelihood of this increases if the patient is admitted to hos-

pital and his eyedrops are forgotten (Fig. 6). A great deal of research with

Fig. 6. Rejection after gall bladder operation (patient had not used eyedrops in hospital).

278

animal experiments has been performed on the rejection of corneal grafts.

Little has been published about the morphology of rejected human grafts.

If the rejection is found to be irreversible the decision to reoperate is taken

(Fig. 7). When this operation is performed the discarded corneal disc can be

Fig. 7. Irreversible rejection.

used for morphological examination. It is obvious that we can only examine

the final stage of the process in this way. Patients with an irreversible rejec-

tion are placed on the waiting list for an HLA-matched graft. It may be

some time before a good match is available so that the interval between the start of the rejection and reoperation varies considerably.

Six cases are described, chosen from 40 cases of irreversibly rejected grafts. The clinical picture in these 6 cases was perfectly typical; in all cases a Khodadoust line was more or less clearly seen.

A short clinical description is given of the cases whose morphology is stu- died in the following article, in order to be able to compare the microscopi-

cal findings with the case history.

W.G. Q 45 years.

Leucoma corneae + congenital cataract on both eyes.

8-10-1975 Perforating keratoplasty OD + cataract extraction.

6-I0-I 976 Perforating keratoplasty OS + cataract extraction.

279

6- 7-1977 14- 7-1977

17-11-1977

Reversible rejection OS (cleared with local corticosteroids). Irreversible rejection OD. The rejection was marked by clouding, ciliary injection, aqueous flare and Descemet pre- cipitates. The clouding continued after the first signs of re- jection and did not improve on local steroid therapy. No vascularisation. Both grafts were totally clear before the rejections. Regrafting OD with HLA-matched donor.

T.T. d 25 years.

Member of a family with van der Hoeve's syndrome; blue sclerae, high myopia, keratoconus, brittle bones, deafness.

15-7-1975

15-6-1977

Perforating keratoplasty. The graft was successful with good vision. After 4 weeks a classical clinical rejection appeared, although this cornea was not vascularized before operation. A clinical rejection occur- red, with ciliary injection, Descemet precipitates, aqueous flare and an endothelial rejection line. In the months follow- ing the first signs of rejection, moderate vascularisation of cornea and graft occurred. Regrafting with tissue-typed donor-material.

W.V. d 45 years.

Herpes Keratitis.

17- 8-1977

15-12-1977

20- 1-1978

Perforating keratoplasty h chaud. Ruptured descemetoc~le, caused by keratitis herpetica. Regrafting. The previous graft melted away and again a kera- toplasty h chaud is needed. Within 4 weeks this graft underwent a clinical rejection, with heavy vascularisation of the cornea. The rejection started with ciliary injection, redness of the eye, aqueous flare and an endothelial rejection line. In spite of local corticosteroids the clouding proceeded.

F.D. d 25 years.

Clouded macula corneae of unknown origin, probably endothelial dystrophy.

280

19-12-1974

14-02-1975

12-02-1978

Perforating keratoplasty.

Irreversible rejection.

The patient visited the clinic two weeks after the first signs

of clouding. At this time the graft was thickened and clouded. Endothelial details could not be seen. Because the graft was perfectly clear for two months and the eye was not inflamed, the sudden onset of thickening of the cornea, ciliary injection and in i ta t ion led us to the diagnosis

of rejection. At operation the cornea and graft were moderately vasculariz- ed. Local corticosteroids did not improve the process.

Regrafting. A double Descemet's membrane was found at

this operation.

G.v.W. 9 32 years.

Traffic accident. Perforating injury by glass of both eyes.

OD was enucleated. OS was saved, but with a cataract and severe scarring of the cornea.

Episodes of high tension. Cataract and glaucoma surgery.

24-01-1974

20-03-1974 01-03-1974

08-04-1978

Perforating keratoplasty OS.

Full vision with correction. Clear graft. Rejection, starting with clouding of the epithelium. Later Descemet precipitates were seen, as well as a rejection

line in the endothelium. Soon clouding of all layers of the graft. The graft did not clear with conservative therapy (local

corticosteroids). The graft was moderately vascularized at the end.

Regraft with tissue-typed and matched graft with good results (up to 1-1-1980).

v .d .H, d 64 years.

Recurrent herpes keratitis.

19-2-1972 Perforating keratoplasty. The graft was clear, but we saw several episodes of reversible

rejection. With each rejection episode, ciliary injection, aqueous flare, Descemet precipitates and occasionally an en-

281

dothelial rejection line were seen.

These episodes ended in an irreversible clouded cornea, with

much vascularisation.

6-4-1978 Clouded graft, irreversible rejection.

21-4-1978 Regrafting with HLA-typed donor.

REFERENCES

Chandler, J.W. & Kaufman, H.E. Graft reactions after keratoplasty for keratoconus. Amer. J. Ophthal. 77:543-547 (1974).

Inomata, H., Smelser, G.K. & Polack, F.M. Fine structure of regenerating endothelium and Descemet's membrane in normal and rejecting corneal grafts. Amer. J. Ophthal. 70:48-64 (1970).

Inomata, H., Smelser, G.K. & Polack, F.M. The fine structural changes in the corneal endothelium during graft rejection. Invest. Ophthal. 9:263-271 (1970).

Khodadoust, A.A. The aUograft rejection reaction; the leading cause of late failure of clinical corneal grafts. Corneal graft failure, ~'ba Foundation Symposium 15: 151- 163 (1973).

Khodadoust, A.A. & Silverstein, A.M. Induction of corneal graft rejection by passive cell transfer. Invest. Ophthal. 15:89-95 (1976).

Maumenee, A.E. The influence of donor-recipient sensitization on corneal grafts.Amer. J. Ophthal. 34:142-152 (1951).

Maumenee, A.E. Clinical aspects of the corneal homograft reaction. Invest. Ophthal. 1:244-252 (1962).

Offret, G., Pouliquen, Y. & Guyot, D. Aspects cliniques des rdactions immunitaires apr6s kdratoplasties transfixiantes chez l'homme. Arch. Ophtal. (Paris) 30:209-218 (1970)

Polack, F.M. Corneal graft rejection: clinico-pathological correlation. Corneal graft fail- ure, Ciba Foundation Symposium 15:127-139 (1973).

Polack, F.M. Corneal transplantation. New York, Grune & Stratton, 1977. p. 227. Sundmacher, R. Immumeaktionen nach Keratoplastik. Klin. Mbl. Augenheilk. 171:

705-722 (1977). Vannas, S., Vannas, A. & Karjalainen, K. HLA-Antigene und AUograftreaktion bei

Keratokonus. Klin. MbL Augenheilk. 170:391-396 (1977).

Authors' addresses: C.C. Kok-van Alphen & H.J.M. V~51ker-Dieben Eye Clinic University Hospital Leiden The Netherlands

J.J.L. van der Want & G. Vrensen The Netherlands Ophthalmic Research Institute Dept. Morphology Amsterdam The Netherlands

282